Merged revisions 81907 via svnmerge from

svn+ssh://pythondev@svn.python.org/python/trunk

........
  r81907 | antoine.pitrou | 2010-06-11 23:42:26 +0200 (ven., 11 juin 2010) | 5 lines

  Issue #8941: decoding big endian UTF-32 data in UCS-2 builds could crash
  the interpreter with characters outside the Basic Multilingual Plane
  (higher than 0x10000).
........

Lib/test/test_codecs.py

@@ -353,6 +353,16 @@ class UTF32Test(ReadTest):
         self.check_state_handling_decode(self.encoding,
                                          "spamspam", self.spambe)
 
+    def test_issue8941(self):
+        # Issue #8941: insufficient result allocation when decoding into
+        # surrogate pairs on UCS-2 builds.
+        encoded_le = b'\xff\xfe\x00\x00' + b'\x00\x00\x01\x00' * 1024
+        self.assertEqual('\U00010000' * 1024,
+                         codecs.utf_32_decode(encoded_le)[0])
+        encoded_be = b'\x00\x00\xfe\xff' + b'\x00\x01\x00\x00' * 1024
+        self.assertEqual('\U00010000' * 1024,
+                         codecs.utf_32_decode(encoded_be)[0])
+
 class UTF32LETest(ReadTest):
     encoding = "utf-32-le"
 
@@ -386,6 +396,13 @@ class UTF32LETest(ReadTest):
         self.assertRaises(UnicodeDecodeError, codecs.utf_32_le_decode,
                           b"\xff", "strict", True)
 
+    def test_issue8941(self):
+        # Issue #8941: insufficient result allocation when decoding into
+        # surrogate pairs on UCS-2 builds.
+        encoded = b'\x00\x00\x01\x00' * 1024
+        self.assertEqual('\U00010000' * 1024,
+                         codecs.utf_32_le_decode(encoded)[0])
+
 class UTF32BETest(ReadTest):
     encoding = "utf-32-be"
 
@@ -419,6 +436,14 @@ class UTF32BETest(ReadTest):
         self.assertRaises(UnicodeDecodeError, codecs.utf_32_be_decode,
                           b"\xff", "strict", True)
 
+    def test_issue8941(self):
+        # Issue #8941: insufficient result allocation when decoding into
+        # surrogate pairs on UCS-2 builds.
+        encoded = b'\x00\x01\x00\x00' * 1024
+        self.assertEqual('\U00010000' * 1024,
+                         codecs.utf_32_be_decode(encoded)[0])
+
+
 class UTF16Test(ReadTest):
     encoding = "utf-16"